Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Support vector machine based white space predictors for cognitive radio

: Al Halaseh, R.
: Hildebrandt, G.; Dahlhaus, D.; Hunziker, T.

Volltext urn:nbn:de:0011-n-1803149 (15 MByte PDF)
MD5 Fingerprint: a6ffa374ffe20bfcd9e65bdd13156ecc
Erstellt am: 20.9.2011

München, 2011, 122 S.
Kassel, Univ., Master Thesis, 2011
Master Thesis, Elektronische Publikation
Fraunhofer ESK ()
channel occupancy; prediction; support vector machine; support vector regression; cognitive Radio

Nowadays, wireless technology has become an essential part of our lives, also in industrial sections. The growing demands on such technologies by time and the newly appeared ones have created a scarce resources problem, due to the limited available resources. Besides, the unlicensed spectrum is fully occupied with users comparing to the licensed one.
Cognitive Radio has emerged with the solution to utilize the available spectrum more efficiently and equally between all users. Its users have the ability to conditionally access both licensed and unlicensed bands without any barriers.
The recent, limited resources approach of employing Support Vector Machine for cognitive radio context is held in this Master thesis, and implemented to predict the channel occupancy. By evaluating its performance using several data sets, this new approach has proved its reliability as a channel occupancy prediction tool.